another WebCGM spec glitch -- p.s.

I wrote:

>[...]I suspect that viewer implementations rely on case-insensitive (and 
>if I were an viewer implementor, I would continue to be forgiving, 
>regardless of what the spec says).

I didn't think this out very carefully.  Maybe a viewer could have a 
special "case-insensitive" mode or option, for dealing with old legacy 
stuff or wrongly written stuff, but it would be dangerous to just ignore 
the rule.

If the spec says case-insensitive, then sooner or later someone (content 
generator) is going to take that literally, even if everyone agrees that 
such constitutes bad practice.  In fact, there should be a test in a future 
release of the test suite (a detailed, or drill-down test, not a basic 
test) which probes this very rule.

...my two cents...

-Lofton.